Characterizing organics

The U.S. Department of Energy’s (DOE’S) Bioenergy Technologies Office (BETO) supports technologies to produce fuels, products and power from non-food sources of biomass and waste resources (organics).

One of BETO’s goals is to push the minimum fuel selling price (MFSP) from $3 gasoline gallon equivalents (gge) to $2.50 gge by 2030. However, this goal will require significant cost reductions in delivered feedstock cost, and a driver of this goal is organics characterization in processing MSW.

BETO is pursuing the study of economically advantaged feedstocks (e.g., MSW, biosolids) beginning in 2022. This implies a need for sufficient quantities of these sources to be researched and studied, as well as the impacts of their intrinsic low quality and heterogeneity to be understood and mitigated.

The numbers best illustrate the need. There are 30 million tons of yard waste produced annually and 13 million tons destined for landfills, according to a 2015 assessment by the EPA. Significantly, there are now organics bans at state levels scheduled to take effect in the next few years, including California; however, adequate capacity is lacking to address such diversion mandates and many conversion technologies require real-time composition analysis to operate properly.

Given the high volumes available and the low costs of these materials, they represent an attractive option to meet BETO’s goals of developing technologies for producing cost-competitive advanced biofuels from non-food biomass resources. Advances in techniques to characterize MSW in real time with respect to content and contaminants are enabling innovation in recycling and environmental protection. The potential of MSW fractions as high-volume, low-cost feedstock to produce biofuels and bioproducts has been recognized but not fully implemented due to knowledge gaps, which include the following: What is the makeup of potential feedstock materials and contaminants in a waste material stream and how do these change with time and location? What is the feedstock potential of materials sampled from a waste material stream? How will feedstock materials with acceptable levels of contamination be separated from waste streams with cost levels and scale sufficient for development? What are techniques to conduct high throughput biofuel characterization determinations? What is the ability to analyze the streams to determine if the field metadata can predict the feedstock potential as a function of time and location?

These knowledge gaps exist not because of the lack of ability to execute the studies; rather, they can be attributed to the fact that the focus in waste characterizations is often limited by funding or deadlines, and narrowly scoped efforts in planning new waste handling operations.

Innovation in detection and sampling methods for the waste stream is one of the BETO study focus areas. MSW streams are complex, and manual separation and characterization are time-consuming and limit the ability to analyze the amount of each fraction of the waste stream and their associated chemical and physical properties. New application and standard imaging and spectrographic methods will enable identification of different components of each waste stream as well as contaminants such as glass and metals. New detection methods will also enable sampling that permits a more detailed quantitative characterization of the waste stream. The potential identification technologies include XRF spectroscopy, NIR cameras, 3D vision capabilities, and visible spectrum (RGB) cameras/sensors.

These technologies have been used to characterize MSW streams to a certain degree, but have not been combined to achieve higher flow rate, identification accuracy, and waste characterization. Additionally, artificial intelligence has application given its ability to collect large and diverse datasets on the waste material stream. AI can identify various components in the stream for a reasonable real-time characterization maximizing the greatest potential for captured data to drive machine-learning results, laying the groundwork for automated extraction of desired materials from the waste stream at an economically advantaged scale in which MSW is collected.